Incremental maintenance of an XML index on binary XML data

• US 7,739,251 B2
• Filed: 03/07/2007
• Issued: 06/15/2010
• Est. Priority Date: 10/20/2006
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Gupta, Nitin, Idicula, Sam, Chandrasekar, Sivasankaran, Agarwal, Nipun
• Primary Examiner(s): Mofiz; Apu M
• Assistant Examiner(s): GIRMA, ANTENEH B
• Application Number:
  US11/715,603
• Publication Number:
  US 20080098020A1
• Time in Patent Office: 1,196 Days
• Field of Search: None
• US Class Current: 707/696
• CPC Class Codes: G06F 16/258 Data format conversion from or to a database
• CPC Class Codes: G06F 16/81 Indexing, e.g. XML tags; Data structures therefor; Storage structures

Abstract

Techniques are provided for incrementally maintaining an XML index built to access XML data that is encoded in binary XML form. Rather than delete and reinsert index entries of all the nodes of a modified XML document, only the index entries of the affected nodes are modified. Consequently, the order key values stored in the index may become inconsistent with the current hierarchical locations of the nodes to which the order key values correspond. Techniques are described for resolving the inconsistencies, and for addressing additional problems that result when the XML index is path-subsetted.

Claims

1. A machine-executed method for incrementally maintaining an index that indexes Extensible Markup Language (XML) data, the method comprising:

identifying a node, within the XML data, that is to be deleted in response to a Data Manipulation Language (DML) operation;

generating a hybrid key for the node;

wherein the hybrid key includes (a) a first set of one or more index order key components and (b) a first set of one or more canonical order key components;

wherein the first set of index order key components reflects a portion of an index order key by which the node is indexed in the index;

wherein the first set of canonical order key components reflects the actual position of the node within the XML data;

converting the hybrid key to the index order key;

wherein the index order key includes a second set of one or more index order key components and a third set of one or more index order key components;

wherein the second set of index order key components equals the first set of index order key components;

wherein converting the hybrid key to an index order key comprises converting the first set of canonical order key components to the third set of index order key components;

wherein the first set of canonical order key components differs in value from the third set of index order key components;

using the index order key to locate a set of one or more index entries within the index;

wherein one of the set of one or more index entries indexes the node; and

deleting the set of index entries from the index.

2. The method of claim 1 wherein:

the index is a path-subsetted index;

the step of converting the hybrid key to an index order key includes converting canonical components of the hybrid key to path-subsetted components; and

using the path-subsetted components to determine the index order key that corresponds to the hybrid key.

3. The method of claim 2 wherein the step of converting canonical components of the hybrid key to path-subsetted components includes determining how many sibling components (a) precede a canonical component, and (b) correspond to nodes that are not included in the index.

4. The method of claim 1 wherein:

the node belongs to an Extensible Markup Language (XML) document that is indexed by said index; and

the set of index entries that are deleted from the index does not include index entries for any nodes of the XML document other than the node and any nodes that descend from the node.

5. The method of claim 1 further comprising:

maintaining, as part of the index, a descendant index that is built on index order key values; and

traversing the descendant index, based on the index order key, to identify the set of index entries to delete from the index.

6. The method of claim 1 wherein the DML operation is a delete operation.

7. The method of claim 1 wherein generating a hybrid key for the node includes obtaining the first set of one or more index order key components of the hybrid key from the index, and obtaining the first set of one or more canonical order key components without using the index.

8. The method of claim 7 wherein the step of obtaining the second set of one or more canonical order key components includes using streaming evaluation to generate the second set of one or more canonical order key components.

9. The method of claim 7 wherein the step of converting the hybrid key to the index order key includes:

a) initially establishing a current order key to include the first set of one or more index order key components;

b) selecting, as a current component, a component of the second set of one or more canonical order key components;

c) determining all immediate children of the current order key;

d) ranking the immediate children of the current order key;

e) identifying the child of the current order key that has a rank equal to the current component;

f) appending to the current order key a component associated with the child identified in step (e); and

g) repeating steps (b) to (f) until all components in the second set of canonical order key components have been selected as the current component.

10. The method of claim 9 further comprising using a descendant index to determine all immediate children of the current order key.

11. The method of claim 1 wherein:

the node belongs to an Extensible Markup Language (XML) document;

the XML document is encoded in binary XML format; and

the DML operation is performed on the XML data using a piece-wise operation that does not cause the XML document to be entirely rewritten.

12. A machine-executed method for incrementally maintaining an index that indexes Extensible Markup Language (XML) data, the method comprising:

identifying a particular node that is to be added to the XML data in response to a Data Manipulation Language (DML) operation;

generating a hybrid key for a parent node of the particular node;

wherein the hybrid key includes (a) a first set of one or more index order key components and (b) a first set of one or more canonical order key components;

wherein the first set of index order key components reflects a portion of an index order key by which the parent node is indexed in the index;

wherein the first set of canonical order key components reflects the actual position of the parent node within the XML data;

converting the hybrid key to a first index order key;

wherein the index order key includes a second set of one or more index order key components and a third set of one or more index order key components;

wherein the second set of index order key components equals the first set of index order key components;

wherein converting the hybrid key to the first index order key comprises converting the first set of canonical order key components to the third set of index order key components;

wherein the first set of canonical order key components differs in value from the third set of index order key components;

adding a component to the first index order key to generate a second index order key for the particular node; and

storing, within the index, an index entry for the particular node;

wherein the index entry includes the second index order key.

13. The method of claim 12 wherein further comprising generating the component to add to the first index order key by:

identifying a last child node of the parent node;

identifying a last component of an index order key of the last child node; and

generating a value, to be used as said component, that is greater than the last component of the index order key of the last child node.

14. The method of claim 12 wherein further comprising generating the component to add to the first index order key by:

identifying two sibling nodes between which the particular node is to logically reside;

identifying last components of index order keys of the two sibling nodes; and

generating a value, to be used as said component, that is between the last components of the index order keys of the two sibling nodes.

15. The method of claim 12 wherein:

the index is a path-subsetted index;

the step of converting the hybrid key to an index order key includes converting canonical components of the hybrid key to path-subsetted components; and

using the path-subsetted components to determine the index order key that corresponds to the hybrid key.

16. The method of claim 15 wherein the step of converting canonical components of the hybrid key to path-subsetted components includes determining how many sibling components (a) precede a canonical component, and (b) correspond to nodes that are not included in the index.

17. The method of claim 12 wherein generating a hybrid key for the node includes obtaining a first set of one or more components of the hybrid key from the index, and obtaining a second set of one or more canonical components without using the index.

18. The method of claim 17 wherein the step of obtaining the second set of one or more canonical components includes using streaming evaluation to generate the second set of one or more canonical components.

19. The method of claim 17 wherein the step of converting the hybrid key to the first index order key includes:

a) initially establishing a current order key to include the first set of one or more components;

b) selecting, as a current component, a component of the second set of one or more components;

c) determining all immediate children of the current order key;

d) ranking the immediate children of the current order key;

e) identifying the child of the current order key that has a rank equal to the current component;

f) appending to the current order key a component associated with the child identified in step (e); and

g) repeating steps (b) to (f) until all components in the second set of components have been selected as the current component.

20. The method of claim 19 further comprising using a descendant index to determine all immediate children of the current order key.

21. The method of claim 12 wherein:

the node belongs to an XML document;

the XML document is encoded in binary XML format; and

the DML operation is performed on the XML data using a piece-wise operation that does not cause the XML document to be entirely rewritten.

22. A computer-readable volatile or non-volatile storage medium storing computer-executable instructions for:

identifying a node, within the XML data, that is to be deleted in response to a Data Manipulation Language (DML) operation;

generating a hybrid key for the node;

wherein the hybrid key includes (a) a first set of one or more index order key components and (b) a first set of one or more canonical order key components;

wherein the first set of index order key components reflects a portion of an index order key by which the node is indexed in the index;

wherein the first set of canonical order key components reflects the actual position of the node within the XML data;

converting the hybrid key to the index order key;

wherein the index order key includes a second set of one or more index order key components and a third set of one or more index order key components;

wherein the second set of index order key components equals the first set of index order key components;

wherein converting the hybrid key to an index order key comprises converting the first set of canonical order key components to the third set of index order key components;

wherein the first set of canonical order key components differs in value from the third set of index order key components;

using the index order key to locate a set of one or more index entries within the index;

wherein one of the set of one or more index entries indexes the node; and

deleting the set of index entries from the index.

23. The computer-readable volatile or non-volatile storage medium of claim 22, wherein the index is a path-subsetted index, and wherein the instructions for converting the hybrid key to an index order key include instructions for:

converting canonical components of the hybrid key to path-subsetted components; and

using the path-subsetted components to determine the index order key that corresponds to the hybrid key.

24. The computer-readable volatile or non-volatile storage medium of claim 23, wherein the instructions for converting canonical components of the hybrid key to path-subsetted components include instructions for determining how many sibling components (a) precede a canonical component, and (b) correspond to nodes that are not included in the index.

25. The computer-readable volatile or non-volatile storage medium of claim 22, wherein:

the node belongs to an Extensible Markup Language (XML) document that is indexed by said index; and

the set of index entries that are deleted from the index does not include index entries for any nodes of the XML document other than the node and any nodes that descend from the node.

26. The computer-readable volatile or non-volatile storage medium of claim 22, further comprising instructions for:

maintaining, as part of the index, a descendant index that is built on index order key values; and

traversing the descendant index, based on the index order key, to identify the set of index entries to delete from the index.

27. The computer-readable volatile or non-volatile storage medium of claim 22, wherein the DML operation is a delete operation.

28. The computer-readable volatile or non-volatile storage medium of claim 22, wherein the instructions for generating a hybrid key for the node include instructions for:

obtaining the first set of one or more index order key components of the hybrid key from the index, and

obtaining the first set of one or more canonical order key components without using the index.

29. The computer-readable volatile or non-volatile storage medium of claim 28, wherein the instructions for obtaining the second set of one or more canonical order key components include instructions for using streaming evaluation to generate the second set of one or more canonical order key components.

30. The computer-readable volatile or non-volatile storage medium of claim 28, wherein the instructions for converting the hybrid key to the index order key include instructions for:

a) initially establishing a current order key to include the first set of one or more index order key components;

b) selecting, as a current component, a component of the second set of one or more canonical order key components;

c) determining all immediate children of the current order key;

d) ranking the immediate children of the current order key;

e) identifying the child of the current order key that has a rank equal to the current component;

f) appending to the current order key a component associated with the child identified in (e); and

g) repeating (b) to (f) until all components in the second set of canonical order key components have been selected as the current component.

31. The computer-readable volatile or non-volatile storage medium of claim 30, further comprising instructions for using a descendant index to determine all immediate children of the current order key.

32. The computer-readable volatile or non-volatile storage medium of claim 22, wherein:

the node belongs to an Extensible Markup Language (XML) document;

the XML document is encoded in binary XML format; and

the DML operation is performed on the XML data using a piece-wise operation that does not cause the XML document to be entirely rewritten.

33. A computer-readable volatile or non-volatile storage medium storing computer-executable instructions for:

identifying a particular node that is to be added to the XML data in response to a Data Manipulation Language (DML) operation;

generating a hybrid key for a parent node of the particular node;

wherein the hybrid key includes (a) a first set of one or more index order key components and (b) a first set of one or more canonical order key components;

wherein the first set of index order key components reflects a portion of an index order key by which the parent node is indexed in the index;

wherein the first set of canonical order key components reflects the actual position of the parent node within the XML data;

converting the hybrid key to a first index order key;

wherein the index order key includes a second set of one or more index order key components and a third set of one or more index order key components;

wherein the second set of index order key components equals the first set of index order key components;

wherein converting the hybrid key to the first index order key comprises converting the first set of canonical order key components to the third set of index order key components;

wherein the first set of canonical order key components differs in value from the third set of index order key components;

adding a component to the first index order key to generate a second index order key for the particular node; and

storing, within the index, an index entry for the particular node;

wherein the index entry includes the second index order key.

34. The computer-readable volatile or non-volatile storage medium of claim 33, wherein the instructions include instructions for generating the component to add to the first index order key by:

identifying a last child node of the parent node;

identifying a last component of an index order key of the last child node; and

generating a value, to be used as said component, that is greater than the last component of the index order key of the last child node.

35. The computer-readable volatile or non-volatile storage medium of claim 33, wherein the instructions include instructions for generating the component to add to the first index order key by:

identifying two sibling nodes between which the particular node is to logically reside;

identifying last components of index order keys of the two sibling nodes; and

generating a value, to be used as said component, that is between the last components of the index order keys of the two sibling nodes.

36. The computer-readable volatile or non-volatile storage medium of claim 33, wherein the index is a path-subsetted index, and wherein the instructions for converting the hybrid key to an index order key include instructions for:

converting canonical components of the hybrid key to path-subsetted components; and

using the path-subsetted components to determine the index order key that corresponds to the hybrid key.

37. The computer-readable volatile or non-volatile storage medium of claim 33, wherein the instructions for converting canonical components of the hybrid key to path-subsetted components include instructions for determining how many sibling components (a) precede a canonical component, and (b) correspond to nodes that are not included in the index.

38. The computer-readable volatile or non-volatile storage medium of claim 33, wherein the instructions for generating a hybrid key for the node include instructions for:

obtaining a first set of one or more components of the hybrid key from the index, and obtaining a second set of one or more canonical components without using the index.

39. The computer-readable volatile or non-volatile storage medium of claim 38, wherein the instructions for obtaining the second set of one or more canonical components include instructions for using streaming evaluation to generate the second set of one or more canonical components.

40. The computer-readable volatile or non-volatile storage medium of claim 38, wherein the instructions for converting the hybrid key to the first index order key include instructions for:

a) initially establishing a current order key to include the first set of one or more components;

b) selecting, as a current component, a component of the second set of one or more components;

c) determining all immediate children of the current order key;

d) ranking the immediate children of the current order key;

e) identifying the child of the current order key that has a rank equal to the current component;

f) appending to the current order key a component associated with the child identified in (e); and

g) repeating (b) to (f) until all components in the second set of components have been selected as the current component.

41. The computer-readable volatile or non-volatile storage medium of claim 40, further comprising instructions for using a descendant index to determine all immediate children of the current order key.

42. The computer-readable volatile or non-volatile storage medium of claim 33, wherein:

the node belongs to an XML document;

the XML document is encoded in binary XML format; and

the DML operation is performed on the XML data using a piece-wise operation that does not cause the XML document to be entirely rewritten.